Supercritical fluids can be used as solvents in extraction instruments, chromatographs and other related instruments. The critical temperature is that temperature above which the distinction between gases and liquids disappears; there is one fluid phase for all pressures. No matter how much pressure is applied, a liquid phase cannot be condensed. The supercritical region is defined by all temperatures and pressures above the critical temperature and pressure. Supercritical fluids are a useful hybrid of gases and liquids as we commonly perceive them, possessing gas-like viscosities, liquid-like densities, and diffusivities greater than typical liquid solvents. The liquid-like density of a supercritical fluid imparts a variable liquid-like solvent power by an essentially linear function of density. This allows the solvent power, usually considered a chemical interaction, to be set ("dialed in") simply by adjusting a physical parameter, namely density or pressure. Variable solvent power is most pronounced over densities corresponding to pressures from about 0.8 times the critical pressure to pressures at which the solvent densities become nearly liquid-like.
Another important aspect of supercritical fluids is that compared to typical liquid solvents the supercritical fluid transport properties of viscosity and diffusivity allow enhanced mass transport within complex matrices, such as coal, plant or animal tissue, or packed beds. In other words, supercritical fluids penetrate better and dissolve almost as well as typical liquids. Therefore, supercritical fluids are more efficient to use for extractions of complex matrices.
Even though the solvent power of a supercritical fluid is variable, each supercritical fluid has a maximum solvent power approaching that of the substance as a liquid. Although the solvent power of a liquid is also variable and also essentially a linear function of the density, far larger increases in pressure are necessary to produce significant increases in liquid solvent power. For a typical laboratory liquid solvent such as methanol, the pressure can be lowered to atmospheric and a solution of solvent and solute is stable. Heat must be supplied to drive off solvent for solution concentration or solute drying. If a liquid with a very high vapor pressure at ambient temperatures is used, such as liquid carbon dioxide or ammonia, the solvent is readily evaporated upon solution decompression, leaving a dry, concentrated solute. However, there is a discontinuity in the density along an isotherm corresponding to tie lines between coexisting liquid and gaseous phases. Because of the discontinuity in density and solvent power essentially is made high or low, "go" or "no-go", and most of the versatility in solvent power control is lost.
In terms of mass transport properties, diffusivity is an important parameter. Liquid diffusivities cover a large range, some of which (e.g., carbon dioxide and ammonia) are almost the same as those of the supercritical fluid when the liquid is at near critical conditions. In those cases either the liquid or the supercritical fluid would suffice for efficient extraction. However, for commonly used liquids, the diffusivities are lower by a factor of 10 to 100 than for supercritical fluids and therefore these liquids are less efficient in terms of time for extraction.
Supercritical solvents are therefore superior in that they (1) allow a large continuous range of solvent powers, thus providing for selective solvation of solutes over that range, (2) provide a means of rapid and complete solvent/solute separation with minimal heat input, and (3) can decrease the time for extraction of complex matrices by a factor of 10 to 100.
Carbon dioxide is the principal extracting fluid used in supercritical fluid extraction systems since it is cheap, innocuous, readily available at high purities, and has a critical temperature of about 31.degree. C. so it is useful for thermally labile compounds. Furthermore, it is mutually soluble with many common liquid solvents. It has been found that carbon dioxide has a solvent power similar to that of hexane. Hence, many applications exist which require great solvent power, the advantageous properties of supercritical fluids, and mild operating temperatures for thermally labile compounds. Mixtures of carbon dioxide plus modifiers can meet these requirements. As well known to those of ordinary skill, supercritical fluids can be used as solvents in extractions and chromatography; in such applications carbon dioxide is the preferred solvent. Other fluids which have critical points near ambient temperature (25.degree. C.) such as ethane, nitrous oxide, ethylene, or sulfur hexafluoride could also function as the base solvent. The capability to utilize these alternative solvents is preferably not exploited because of the inherent danger in using these solvents.
Another way to increase the solute solubility in supercritical fluid systems is through temperature control. It has been demonstrated that for some solutes a certain minimum temperature is necessary for appreciable solute solubility in a supercritical fluid, even at maximum densities, so temperature is also a solvent power parameter beyond the simple P-T relationship with density.
The problem encountered in building extraction systems which use compressible fluids at pressures above atmospheric (ambient) is that the pressure of the fluid system is coupled with the mass flow rate of the flowing fluid. This is a result of the use of restrictors having constant geometry to achieve a pressure drop from higher pressures to lower pressures which are normally but not restricted to ambient or vacuum conditions, thus the only practical way to achieve higher pressures upstream of the restriction is to increase the mass flow rate. However, in operating these extraction systems it is highly desirable to have the same mass flow rate at all pressures, i.e., with the lower flow rates to achieve lower pressures, the time to flow a chosen net amount of fluid can be significantly longer than at higher pressures, e.g., hours vs. minutes. Therefore, in a system which can be adjusted to operate sequentially at multiple pressures (e.g., 1100 psi, 1300 psi, 1500 psi, 3000 psi, 6000 psi) at flow rates which provide the same integrated mass of expanded carrier fluid at each pressure setting the time to complete the method can be so long that the method becomes inefficient. This is particularly true in operating supercritical fluid extraction systems, as well as near-critical fluid extraction, which can be more efficient in time than extractions using conventional fluids, due to the transport properties (e.g., viscosity, diffusivity) of critical and near-critical fluids. However, that efficiency is lost when the parameters of pressure and flow cannot be decoupled because of the hardware implementation, e.g., constant geometry flow restrictions.
It is therefore an object of the present invention to decouple the pressure and flow parameters so that they can be set and controlled independently.
In the past, operators of supercritical fluid chromatograph (SFC) and supercritical fluid extractor (SFE) instruments selected pressure and temperature parameters to set the desired density parameter indirectly. Density is usually a more meaningful parameter than pressure in supercritical (and near-critical) systems due to its simple linear relationship to the fluid's solvent power. Often, in any previous implementations of supercritical fluid systems, only pressure and temperature inputs have been allowed; calculating the resultant density or even understanding the pressure/density/temperature relationship has been left to the user, albeit occasionally prompted by abbreviated look-up tables. Therefore, in those cases, controlling the solvency has relied upon the technical sophistication and persistence of the user to apply equations of state appropriately.
The extraction of solids has typically been done manually in conventional laboratory glassware. Laboratory devices which automate the steeping of the sample in a boiling liquid solvent along with the subsequent concentration of the resultant solution are sold commercially as the "Soxtec" by Tecator. This device does not permit fractionation as part of the method; the output is not in autosampler-compatible vessels and samples are boiled at temperatures determined by the boiling points of the extracting solvents, which can be detrimental to thermally labile compounds. Recently, several companies have offered instruments based on using supercritical fluids: Milton Roy, Suprex, CCS, Lee Scientific/Dionex, and Jasco. The implementations by these companies are not exceptionally sophisticated or automated; they require various degrees of manual intervention. In these implementations, the full capability of using supercritical fluids for quantitative extraction is not realized.
Thus, there exists a long-felt need to design an instrument which could be used to isolate various materials from solid samples using supercritical fluid technology. Either the isolated materials (extract) or the material remaining (raffinate) could be of interest. The process of isolation is referred to as extraction. The need also existed to automate the extraction of solids, since typically the procedures are manually intensive. In the business of preparing samples for analytical instruments, it would be highly desirable to automate sample preparation of raw samples to obtain a form compatible with introduction into typical analytical instruments used in analytical laboratories worldwide. The extraction of solids is often accompanied by other manipulations such as fractionation (e.g., by column chromatography), concentration, solvent exchange, and reconstitution. A method comprised of these generic manipulations may actually consist of many (ranging from several to hundreds) of manual steps. It would therefore be desirable to produce an instrument which replaced the five mentioned generic manipulations so that the only step required was to present a sample to the extraction instrument and receive fractions of isolated material in vessels compatible with the autosamplers of analytical instruments.
Improvements in the isolation of selected material from solid samples and the automation of the generic manipulations of processes such as extraction, fractionation, concentration, solvent exchange, and reconstitution are clearly desirable goals. It would be desirable and of considerable advantage to use a technique different from the traditional laboratory manipulations known to those of ordinary skill if such a different technique provided superior performance in any of the following areas: quantitation, repeatability and reproduciblity, speed, automation, automatability, health hazard reduction, or cost reduction (materials and labor).
Ideally, an extraction instrument would be a stand alone module, that is, a component instrument within an integrated system. In such systems it is also highly desirable to provide robotic accessibility for inputting the sample and retrieving the fraction in an automated fashion. An important aspect of any extraction system therefore lies in its upgradability for automating the sample input (i.e., adding an autosampler for solid samples) and automating the fraction output (i.e., fraction collection with further chemical manipulation and largescale queueing); an external pump module for blending extraction fluid composition could also be a future addition.
Another improvement in the utility of extraction systems would be the use of standard input and output containers for the sample which is input and the fraction solutions which are output. The instrument would accordingly be provided with a multiple sized extraction chamber which accepts the solid samples.
An improved extraction instrument would also incorporate an automated "method", with a method being defined as that sequence of manipulations associated with any one sample. Examples include, but are not limited to, valve actuation, thermal zone setpoint changes, extraction fluid solvent selection, reconstitution solvent selection, flow path selection, fraction output destination, density/pressure setpoint changes, flow rate setpoint changes. Such an automated extraction instrument would also provide automatic selectable output destinations within a method, and from method to method; automatic selectable inputted extraction fluid composition within a method and from method to method; and automatic selectable rinse solvents for the solvent exchange/reconstitution emulation processes within a method and from method to method.